Run Your First Grading Workflow

This quickstart walks you through one complete grading cycle:

1. start the API
2. create a grading configuration
3. submit student code
4. fetch the graded result

Prerequisites

• Docker and Docker Compose installed
• Python available on your machine
• Project dependencies installed (pip install -r requirements.txt)

1) Start Autograder

make start-autograder

When ready, the API is available at http://localhost:8080.

You can check health:

curl http://localhost:8080/api/v1/health

2) Create a grading configuration

curl -X POST http://localhost:8080/api/v1/configs \
  -H "Content-Type: application/json" \
  -d '{
    "external_assignment_id": "demo-assignment-1",
    "template_name": "input_output",
    "criteria_config": {
      "base": {
        "weight": 100,
        "subjects": [
          {
            "subject_name": "Functionality",
            "weight": 100,
            "tests": [
              {
                "name": "expect_output",
                "weight": 100,
                "parameters": {
                  "inputs": ["Alice"],
                  "expected_output": "Hello, Alice!",
                  "program_command": "python3 main.py"
                }
              }
            ]
          }
        ]
      }
    },
    "languages": ["python"],
    "setup_config": {
      "required_files": ["main.py"],
      "setup_commands": []
    },
    "include_feedback": true,
    "feedback_config": {
      "general": {
        "report_title": "Evaluation Report",
        "show_score": true,
        "show_passed_tests": false,
        "add_report_summary": true
      }
    }
  }'

3) Submit a student solution

curl -X POST http://localhost:8080/api/v1/submissions \
  -H "Content-Type: application/json" \
  -d '{
    "external_assignment_id": "demo-assignment-1",
    "external_user_id": "student-1",
    "username": "student_1",
    "language": "python",
    "files": [
      {
        "filename": "main.py",
        "content": "name = input().strip()\nprint(f\"Hello, {name}!\")"
      }
    ]
  }'

Copy the returned id from the response. That is your submission_id.

4) Poll for the result

curl http://localhost:8080/api/v1/submissions/<submission_id>

When grading finishes, response status becomes completed, and you will get:

• final_score
• feedback
• result_tree
• focus

What you just exercised

You ran the full core flow:

LOAD_TEMPLATE -> BUILD_TREE -> SANDBOX -> PRE_FLIGHT -> GRADE -> FOCUS -> FEEDBACK

To understand each stage in depth, continue with Pipeline Architecture.